Infectious disease prevention studies often aim to test or estimate the "causal effect" of a preventive measure on outcomes by comparing the potential outcomes individuals would have under treatment versus control. Examples of outcomes of interest include infection incidence or post-infection outcomes (e.g., disease severity, death). Two analytical challenges of interest exist for these studies. First, analyses on post-infection outcomes are subject to selection bias as only a subset of the randomized population become infected (i.e., infection status is a post-randomization measure on which analyses are often conditioned). Treatment comparisons conditional on post-randomization measures using standard analytic methods do not have a causal interpretation in that the estimates obtained are not unbiased estimates of the contrast between potential outcomes. The principal stratification framework provides estimates of treatment causal effect in the presence of potential selection bias due to post-randomization measures; however, existing methods comprise only Bayesian or large-sample frequentist approaches. To date a general approach to randomization inference within principal strata has not been developed. Furthermore, while principal stratification approaches are abundant in statistical literature, their presence as an applied analytic approach within infectious disease journals is limited. The second challenge of prevention studies involves the analysis of repeated low-dose mucosal challenge preclinical studies of potential vaccines which are becoming more prevalent in an attempt to conduct studies that better mirror 'real life' human transmission. Current statistical literature exploring the analysis of these studies is somewhat limited and simulation results for certain proposed analytic approaches have demonstrated an inflated type I error. Therefore, in this dissertation we 1) develop methods for exact randomization-based causal inference within principal strata in the presence selection bias due to post-randomization measures, 2) present a discussion of selection bias in randomized studies and the use of principal stratification analytic approaches for handling such bias targeted at subject-matter investigators and 3) present a discussion of appropriate analytic approaches for repeated low-dose challenge preclinical vaccine studies.
|Advisor:||Hudgens, Michael G.|
|Commitee:||Cole, Stephen, Koch, Gary, Sen, Pranab K., Wallace, Dennis|
|School:||The University of North Carolina at Chapel Hill|
|School Location:||United States -- North Carolina|
|Source:||DAI-B 73/11(E), Dissertation Abstracts International|
|Keywords:||Causal inference, Exact p-value, HIV prevention, Permutation test, Selection bias|
Copyright in each Dissertation and Thesis is retained by the author. All Rights Reserved
The supplemental file or files you are about to download were provided to ProQuest by the author as part of a
dissertation or thesis. The supplemental files are provided "AS IS" without warranty. ProQuest is not responsible for the
content, format or impact on the supplemental file(s) on our system. in some cases, the file type may be unknown or
may be a .exe file. We recommend caution as you open such files.
Copyright of the original materials contained in the supplemental file is retained by the author and your access to the
supplemental files is subject to the ProQuest Terms and Conditions of use.
Depending on the size of the file(s) you are downloading, the system may take some time to download them. Please be